Experimental Determination of Air/Water Partition Coefficients for 21 Per- and Polyfluoroalkyl Substances Reveals Variable Performance of Property Prediction Models

Abstract

Per- and polyfluoroalkyl substances (PFAS) are a group of chemicals of high environmental concern. However, reliable data for the air/water partition coefficients (K aw), which are required for fate, exposure, and risk analysis, are available for only a few PFAS. In this study, K aw values at 25 °C were determined for 21 neutral PFAS by using the hexadecane/air/water thermodynamic cycle. Hexadecane/water partition coefficients (K Hxd/w) were measured with batch partition, shared-headspace, and/or modified variable phase ratio headspace methods and were divided by hexadecane/air partition coefficients (K Hxd/air) to obtain K aw values over 7 orders of magnitude (10–4.9 to 102.3). Comparison to predicted K aw values by four models showed that the quantum chemically based COSMOtherm model stood out for accuracy with a root-mean-squared error (RMSE) of 0.42 log units, as compared to HenryWin, OPERA, and the linear solvation energy relationship with predicted descriptors (RMSE, 1.28–2.23). The results indicate the advantage of a theoretical model over empirical models for a data-poor class like PFAS and the importance of experimentally filling data gaps in the chemical domain of environmental interest. K aw values for 222 neutral (or neutral species of) PFAS were predicted using COSMOtherm as current best estimates for practical and regulatory use.

Intermolecular Interactions, Solute Descriptors, and Partition Properties of Neutral Per- and Polyfluoroalkyl Substances (PFAS)

“…In a previous article, it was shown that 1H,1H,7H-perfluoroheptyl iodide (6:1 FTI-7H) has a relatively high affinity for water compared to 6:1 FTI. The electron-withdrawing F atoms were thought to make the H atom at the terminal methyl group of 6:1 FTI-7H significantly H-bond donating . In fact, the A of 6:1 FTI-7H is positive (0.24) and higher than that of 6:1 FTI (0.13).…”

“…Thus, the left side of eq was known, and MLR was performed to obtain S , A , and B on the right side. Log ⁡ K false( or log nobreak0em.25em⁡ k false) − c − v V − l L = s S + a A + b B To determine the three descriptors, a minimum of three k and K data are required for a given PFAS compound. The data used here were: k measured for 60 PFAS in this study, K Hxd/w for 21 PFAS, K ow for 12 PFAS (see the Results and Discussion section), and K oa for 3 PFAS . Note that K aw for 21 PFAS in ref was not considered here to avoid duplicate data use, as these data were derived from K Hxd/w and K Hxd/air through the thermodynamic cycle calculation.…”

“…The data used here were: k measured for 60 PFAS in this study, K Hxd/w for 21 PFAS, K ow for 12 PFAS (see the Results and Discussion section), and K oa for 3 PFAS . Note that K aw for 21 PFAS in ref was not considered here to avoid duplicate data use, as these data were derived from K Hxd/w and K Hxd/air through the thermodynamic cycle calculation.…”

“…The programs Turbomole, COSMO ConfX , and COSMO thermX (version 2021, COSMO logic , Dassault Systèmes) were used for the calculation. More details about the method have been explained elsewhere. , …”

“…Even basic property data such as octanol/water partition coefficients ( K ow ) are scarce for PFAS. − To the best of my knowledge, only one scientific paper has reported measured K ow values for neutral PFAS with perfluoroalkyl chain lengths of 4 or more. Similarly, reliable air/water partition coefficients ( K aw ) were available for only a few neutral PFAS with (CF 2 ) ≥4 until our group recently reported experimental values for 21 PFAS …”